BOOTSTRAP HYPOTHESIS TESTS FOR EVOLUTIONARY TREES AND OTHER DENDROGRAMS

The bootstrap computer-intensive statistical technique is frequently a pplied to statistical analyses of phylogenetic trees. The widely used rule that a group is supported significantly if it appears in at least 95% of bootstrap trees is conservative in most situations. This paper describes three ways of using the bootstrap to carry out statistical inference on phylogenies. The first method tests whether there is nonr andom support for a single group or tree. The second method compares t he support for two groups or trees. The third method tests whether a s ingle group or tree has better support than the set of all possible al ternatives; this may be a replacement for the ''95% rule.'' These test s generally require fewer bootstrap trees to be estimated than do othe r methods of bootstrapping phylogenies. A simple, sequential statistic al method tan be used to increase the efficiency further. These method s can be applied to tests of multiple hypotheses about a single phytog eny. Parsimony analyses of 5S rRNA sequences of plants and cluster ana lyses of randomly amplified polymorphic DNA bands in three pathotypes of the cereal eyespot fungus are used as illustrative examples. The te sts can be used to analyze dendrograms in subjects other than taxonomy .